View Full Version : Measuring loudspeakers - what do you believe?

06-10-2011, 03:50 PM
There are two broad schools of loudspeaker design: 'rationalists' who are rooted in measurement* and objectivity and the subjectivists who design purely by listening. Most of us fall somewhere in between. We wish that measurements could tell the whole story (if they could, we could design on a laptop on a tropical beach, azure blue sea, endless G&Ts confident that when we sent the design files to Production the speaker would be an immediate success) but they don't. The subjectivists would argue that not only do they not tell the full story they completely fail to describe the listening experience and are, therefore, valueless.

Surely the truth must lie between these extremes. Measurements (frequency response, distortion, dispersion etc.) definitely illuminate some aspect of the design. That is, modern audio analysis computer systems can produce much pretty, multi-coloured even 3D graphical data about the speaker under test. Can we - should we, dare we - trust our eyes when examining these plots? Has an information overload given the designer the illusion that he is more in command of the design than he really is? How is it possible that the BBC was able to produce great loudspeakers in the 1960s entirely 'by hand' a decade before the first calculator? I'd say it was down to careful interpretation of the little measured data that was available combined with a willingness to listen and let the listening experience be the final arbiter - not the formulae and rule book. It may even be that the smaller quantity of difficult and expensive to acquire measured data gives a clearer picture of the speaker's behaviour than the vast quantity of press-of-a-button data that is so readily achievable now. Just as anyone with an internet connection can swat-up and become (their own) brain surgeon, loudspeaker measurement systems are affordable by serious hobbyists; it's a wonder that there are not even more speakers on the market. Factor X that takes a design from the flawed to the great cannot nowadays be related to audio measurement equipment - it must be somehow related to the ability to wade through, discard and interpret the vital, core data. And know what to ignore.

As mentioned previously, we recently relocated Harbeth's R&D facility to a beautiful, spacious 16 century barn. The 500 year old oak beams across the high ceiling have given me the reflection-free heigh to allow me to build an all-weather, fixed, loudspeaker measurement set-up. By manipulating the measurements it's possible to make full-range 20Hz to 20kHz acoustic measurements of any speaker in this ordinary room; something that would have (and did) require a large anechoic chamber before. And I've spent over two months fine tuning the set-up to remove (or at least damp down) reflections so that what the microphone picks-up is as much about the speaker as possible and not the speaker + room. After many long days and evenings fiddling around with positions, angles, heights, path lengths, Rockwool, VetBed wadding and even duvet quilts I have measurements which are as believable as those I took at the BBC anechoic chamber on the same speaker. I can't say more than that. I can't say that there or even the BBC measurements are 100% accurate because there are no absolutes in acoustics. But they at least agree. And that is all one can say. Perhaps I'll write more about the set-up another time.

Objectivist speaker designers place great weight on their frequency response measurements. These, they believe, fundamentally describe how the speaker will perform, and some would work long and hard to achieve a certain shape to the loudness frequency response, plotted on graph paper from the lowest the the highest frequency. Some would perhaps aim for a response that gradually increases in loudness from the lowest to highest frequencies. Another for a basically flat response. Another for a response with a gentle incline from LF to HF. The variables that they'd be playing with would be the latent acoustic characteristics of the drive units (if they actually made the drivers in-house they could play around with the moving parts to their hearts content to achieve the desired response shape) and the crossover/equaliser aka the network.

Now a complex network allows the designer to 'nip and tuck' the system's acoustic output. If he wants to reduce a peak in the speaker's acoustic output (frequency response) he would open an electrical pathway (a controlled, mild short circuit) between the red and black terminals on the speaker box so that less volts arrive at the drive unit in a certain frequency band that it is mechanically rather energetic in - the result would be to flatten an acoustic peak. Filling-in holes in the frequency response is a much more difficult problem because if you could create sound out of no (or too little) sound then the laws of physics would have been beaten. You'd have to reduce the entire audio band by a certain amount to bring the peaks down to the level of the troughs, and in so doing would discard useful sensitivity and introduce crossover complexity, a low system impedance and much cost and size. So in reality, the speaker designer is concerned (or at least, we are) with reducing the peaks, not filling the holes (if there are any), maintaining a good overall sensitivity and high-ish system impedance.

So, when I look at a frequency response curve, I know that my hands are tied concerning filling in any holes or dips in the frequency response by electrical means in the crossover. Those rightly should be addressed in the design of the drive unit - ' a mechanical problem should best be solved mechanically in the drive not electrically in the network' is a good starting position. I'm interested in the peaks, those frequency bands where for often inexplicable reasons the drive units and/or cabinet have extra efficiency, produce more dBs of output. We may be able to bring down those high-energy bands in the crossover/equaliser network.

Attached is a graph of a mini-monitor measured at the BBC chamber and in my new set-up. The red and light blue are basically the same data of the speaker at the BBc chamber in 2008. The yellow is a measurement of the same speaker here, a few days ago. The vertical axis is dB, only 2dB/division (a high resolution display). You'll see that above about 3kHz (horizontal axis) the latest and old measurements are very similar and could have been made even more so had I spent time exactly aligning the microphone. Up to 1kHz the curves are basically the same but between 1 and 3kHz (the presence region) there is about a 2dB difference for the very same speaker. Since the HF (and lf) are consistent over the three years, what we are witnessing is some ageing process in the speaker where the output in that band has increased in loudness. Or is it a measurement error?

Now, back to the point: had I been developing this speaker now in 2011 and had not measured this speaker in 2008, I would have assumed that the peak at 60dB 2kHz was to be pulled down in my network (to, say, the level at about 1kHz, 58dB) and would have thrown away that 2dB in the crossover by using a combination of a coil, a resistor and a capacitor - cost perhaps GBP 7-10 per speaker. But that would reduce the overall efficiency for what is already a low-efficiency speaker system, possibly eaten into the power excursion capability (limited at LF) and only to find ..... that this peak was unstable and moved about in frequency and amplitude as the seasons pass. Hence the danger of failing to take a long, cool, rational view of any speaker measurements.

* When we talk of 'loudspeaker measurement' we mean placing a microphone somewhere near the loudspeaker under test, injecting a test signal into the speaker and plotting a frequency v. measured parameter (loudness, distortion, dispersion) initially on paper chart, now on the computer screen.


06-10-2011, 08:22 PM
Very interesting indeed, Alan.

If you were going into production with this speaker, would you listen to a prototype or two, to determine if the peak was a problem or not before deciding to flatten it with a network ?

Any chance of a pic or two of the new facility, when you get some time ?

07-10-2011, 08:11 AM
... If you were going into production with this speaker, would you listen to a prototype or two, to determine if the peak was a problem or not before deciding to flatten it with a network ?That is a really astute observation Dave. I've been mulling over how to answer it. Ten speaker designer would probably provide ten different answers.

I've tried to step outside my own methodology and to consider your point as if you were asking our founder, Dudley Harwood, how he would approach this issue. I was reminded of a conversation with him when he visited me four or five years after he retired and when we were in production of the LS3/5a, the speaker he designed at the BBC and knew inside out. I enthusiastically showed him the grading and sorting process of the B110/T27 drive units destined for our 3/5as. He was far from impressed and left me floundering for a comment. He said 'why are you going to all the bother of selecting and matching drivers and crossovers to that standard? You've made a rod for your own back ... you've set a frequency response selection spec. which is far in excess of the average human ear's acuity ... slowed down production and added labour cost. Taken together that is nothing to be proud of. In fact over-engineering is bad engineering'.

I believed - and still do - that I was adding value at the same time as justifiably adding cost. Our 'super-selection' above and beyond what the average listener could detect gave us a safety margin so that even when all the variables stacked-up against us we could be confident about what we were making and the performance variation across a batch.

Turning to the design process as you mention, if the loudspeaker is to have a basically neutral character then the designer perhaps should be aiming for a fairly flat frequency response, as a staring point. If not there is a danger that the speaker will sound 'over-blown' in some frequency bands, and recessed in other. But, as you note, how far do you go even if you are the most dedicated, fanatical objectivist with all the time and tricks to flatten every peak in the response curve? As I believe that there are probably listeners who have greater sensitivity and acuity than I do, I take my responsibilities (as Harwood commented) to present as flat a technical response as possible. Maybe that's a mistake. Maybe there aren't stern critic out there who can hear lumps and bumps which I've removed by relatively complex and expensive means. I'll never know. And maybe they aren't audible anyway - and maybe Harwood is right - but I like to see a basically flat response unless there are very good sonic reasons not to. So, if I see an excess of energy in a band of the frequency response plot, my instinct is to attack it (in the driver or network) so that the speaker basically has a straight-line input>output nature.

And yes, that hugely extends the development time, cost and mental effort. But if you are working entirely and totally alone, seeking no input from any other listener until the day the speaker goes to market, I think it's best to err on the side of caution and carefully consider all deviations from the perfect flat-line theoretical speaker. Audible or not to me.

07-10-2011, 04:49 PM
.... But, as you note, how far do you go even if you are the most dedicated, fanatical objectivist with all the time and tricks to flatten every peak in the response curve?...Here is an example of just this problem.

I've highlighted a bump in the frequency response of a loudspeaker. It's one of several that I could have picked. Now, the first issue is to have a think about what the source of this bump actually is. Is it something to do with the drive units, probably the woofer at this mid frequency? Or some oddity of the way the drive units are mounted on the baffle? Or the shape and dimensions of the baffle? Or the grille? Or some internal structural or airborne resonance? Or a crossover fault perhaps? Or something to do with the measurement set-up - a reflection off a nearby object corrupting the measurement? Or some other obscure phenomena. It could in fact be partly attributable to all of these to one degree or another. It could take many hours or even days to find the primary source.

So what do we do? Just because with some effort we can dive into the crossover/equaliser network and clobber this peak by adding 'suck-out' components - should we? Is there a danger that we could unwittingly introduce a fix which whilst producing a nice flat technical response, actually sounds more objectionable than leaving the peak unadjusted? The answer is - yes. We absolutely must identify what is going on - why there is a boost in the output and from where it originates.


07-10-2011, 07:37 PM
In a sophisticated BBC-style crossover we can not only split the incoming sound and send bass and tweeter to the correct drive units, we can adjust the overall speaker's response shape, and as mentioned before, by adding even more components (and cost) tame lumps and bumps in the response curve if that's our objective.

Some of these adjustments can attack a relatively narrow band of frequencies. Other have a much wider sweep that effects a whole band. Take for example the attached mini-monitor I measured today. The grey curve is how the speaker measures when I connect the amplifier normally to the red and black rear terminals on the cabinet. Both woofer and tweeter are being driven. The yellow curve is the result of me opening the cabinet, disconnecting the tweeter, and wiring the woofer directly to the rear terminals, completely bypassing the crossover then carefully sealing the cabinet again.

When we study the yellow 'direct drive' curve we can see several interesting things compared with the frequency response via the crossover:

In the bypass or 'raw' mode the woofer has a gently rising response ...
... which means that it's efficiency is increasing with rising frequency until ...
... about 1kHz when it takes a dive downwards ...
... and drops by about 6dB which means it becomes half as loud by about 1.5kHz, half an octave later
... then stabilises, drops again at about 4kHz (around crossover frequency) ...
... and finally exhibits a huge peak at about 5kHz which means ...
... that this woofer is at its loudest, its most efficient, not in the bass or midrange ...
... but at the bottom of a tweeter's normal operating range i.e. 5kHz or so.

That is a serious issue. If we go for a minimalist crossover with just a component or two (a barmy fad) we will have to accept some or all of the consequences of either a rising response (a subjectively and measured weak bass compared to a prominent mid) and/or a peaky top. If we are trying to turn this 'raw' off-the-shelf drive unit into the bass/midrange driver in a high quality speaker with a tolerably flat frequency response (the grey curve) we are going to have to implement a seriously well engineered crossover with lots of expensive components to smooth out these characteristics. That's going to take design effort and greatly add to the component cost. And who knows how it will sound ....

Which issue are we going to tackle first? Assuming the designer does not make his bass/midrange units (Harbeth does) but buys them in from a commercial company out of a catalogue, take-it-or-leave-it, just where is the poor crossover designer going to start to tackle this remedial design. To turn a sows ear into a silk purse at a cost which the customer will pay for?


07-10-2011, 07:50 PM
I would imagine that given the experience you have gained over many years of tweaking and tailoring, you would be able to fairly quickly knock up a circuit to tailor that rocky mountain curve. Eevn if you could get it only part way there, by listening and measuring you should then be able to ascertain if you are going in the right direction or not. Then, if it sounds promising, you would be able to spend more time getting it just right.

Going back to your previous post which shows a large blip - just how do you go about deciding exactly where such a fault lies. If it was a purely mechanical aberration, you wouldn't want to 'clobber' it with an extra filter. It would be best to locate the mechanical problem and iron that out.

Rather you than me.

Paul G Smith
08-10-2011, 05:32 PM
Wouldn't a mechanical resonance in a drive unit show up on that driver's raw (no crossover) impedance / frequency curve? That would help to see if a peak was due to an acoustic or or mechanical problem.

A small movement in the measuring position might also show if it is a measurement problem. If so then averaging measurements may also be helpful.

10-10-2011, 06:28 PM
Wouldn't a mechanical resonance in a drive unit show up on that driver's raw (no crossover) impedance / frequency curve? That would help to see if a peak was due to an acoustic or or mechanical problem.

A small movement in the measuring position might also show if it is a measurement problem. If so then averaging measurements may also be helpful.All potentially true. Just to explain for others - the reason that you say ....

Wouldn't a mechanical resonance in a drive unit show up on that driver's raw (no crossover) impedance / frequency curve? is because a moving coil loudspeaker is simultaneously a moving coil microphone. The two functions are bonded together and inseparable. So, if the drive units cone is commanded by the music signal to pull inwards the very action of the movement of the voice coil away from its at-rest position (in the close proximity of the driver's magnet) means that a voltage will appear in the voice coils wire in opposition to the music signal. And that means that the drive unit must be acting both as a loudspeaker (genrating sound from incoming voltage) at the same instant that it is acting as a microphone (generating voltage from the motion producing sound). And yes, a close study of the impedance curve (which shows the relative relationship between the voltage pushed into the voice coil from the amp and the voltage generated by the coil in motion) will indeed tell us something about mechanical instability in the drive unit. Reason: if the drive unit has a some sort of undamped resonance at a particular frequency, even after that note has ceased in the music, the diaphragm will bound backwards and forwards ... and that motion will generate a voltage in the voice coil - which will be evidenced as a ripple in the otherwise relatively smooth impedance curve.

An example of this speaker/microphone duality can bee seen in this data sheet from the excellent SEAS company - picked at random here (http://seas.no/index.php?option=com_content&view=article&id=74&Itemid=98). It's not possible without much careful investigation to say whether this would be an audible issue or not.

You are also absolutely right that to take an acoustic measurement at just one point in space would be foolish. The serious designer must discover how the speaker radiates sound all over the listening area.

Where else can you read this stuff I wonder?!

10-10-2011, 06:49 PM
A few post back I illustrated a measurement of a well know mini-monitor showing some peaks in the frequency response and asked if they were 'real' or measurement abberations. And if those peaks (just another way of saying 'louder in those frequency bands') are really there, can we/should we make an effort to tame them, make the speaker less loud and flatten the peaks.

Now that I have, finally, a measurment set-up here in this 500 year old barn that I have confidence in (that is, I believe that the influence of the environment is moderate and that I'm really measuring the speakers themselves) I thought I'd measure a few 1970s mini-monitors that I have to hand. I've mentioned before that speakers, like people and cars, do age and as they do, their technical characteristics slowly change. This particular mini has a feverish following and buyers will pay vast sums for a used pair.

I've overlaid the curves from three (single) mini monitors of the same model and of similar vintage. All three were in use at a BBC regional centre. Would you want to spend big money on one with a huge peak right where the ear is has a sensitivity peak (http://en.wikipedia.org/wiki/Equal-loudness_contour) - around 1kHz?

On another thread here (http://www.harbeth.co.uk/usergroup/showthread.php?1393-Fast-and-slow-speakers-should-fast-painful-to-listen-to) we're commenting on 'fast' and 'slow' speakers - vague terms that we are unable to pin down. But I don't doubt that the mini monitor with the big attention-grabbing peak shown on the attached would, thanks to the ear's sensitivity gain around the peak frequency, sound extremely 'fast' and 'alive'.